Explore the challenges of tracking and ordering events in microservices architectures in this conference talk. Delve into the complexities of synchronizing clocks, understanding the limitations of timestamp-based approaches, and learn about alternative solutions for establishing meaningful order in distributed systems. Discover concepts such as happened-before relation, vector clocks, and correlation IDs. Gain insights into the misconceptions surrounding time in distributed environments and explore mitigation strategies. With examples in .NET, this moderately intense session covers topics including timezones, Daylight Saving Time, leap seconds, Network Time Protocol (NTP), Lamport timestamps, vector clocks, Byzantine generals problem, and consensus algorithms like Raft. Enhance your understanding of log management and event ordering in microservices to improve debugging and troubleshooting capabilities.
Ordering the Chaos - Cleaning Logs and Ordering Events in Microservices
Overview
Syllabus
Intro
About me
What is timezone
Timezones change
Daylight Saving Time
Leap second
How is time distributed
Network Time Protocol (NTP)
Other approaches
Lamport timestamp
Implementation
Logger
Step 1
Step 2
Step 3
Step 4
Vector clock
Byzantine generals
Consensus problem
Raft
References
Taught by
NDC Conferences
